Abstract
Background: Bladder cancer (BLCA) is a common urogenital malignancy with significant heterogeneity, and gemcitabine serves as a key chemotherapeutic agent for BLCA. This study examined gemcitabine sensitivity-related long noncoding RNA (GSRlncRNA) in BLCA and constructed a prognostic signature. Methods: Gene expression and clinical data were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. The gemcitabine sensitivity of BLCA samples was quantitatively calculated. Using weighted gene coexpression network analysis (WGCNA), we analyzed the role of GSRlncRNA in BLCA. We developed a prognostic signature using machine learning algorithm integration, with the TCGA-BLCA cohort (n=405) serving as the training set and GSE31684 (n=93) serving as the validation cohort. Patients were stratified into low- and high-risk groups based on risk scores. Comparative analyses included prognosis, immune infiltration, mutational profiles, and pathway enrichment. Single-cell RNA sequencing (scRNA-seq) and in vitro experiments further validated the expression and functional roles of GSRlncRNA in BLCA. Results: GSRlncRNAs were identified via WGCNA. The partial least squares regression for Cox (plsRcox) was selected as the optimal modeling approach. A prognostic signature comprising 37 GSRlncRNAs effectively stratified patients into distinct risk groups, which demonstrated significant differences in survival outcomes, pathway enrichment, immune infiltration, and mutational profiles. To enhance clinical utility, a nomogram and web-based calculator were developed. scRNA-seq clarified the critical roles of LINC00930 and EMX2OS in BLCA progression. Real-time quantitative polymerase chain reaction confirmed the differential expression of GSRlncRNA in BLCA compared to that in normal tissues. Functional assays demonstrated that LINC00930 and EMX2OS suppression impaired BLCA cell proliferation, migration, invasion, apoptosis, cell cycle, and gemcitabine sensitivity. Conclusions: We identified critical prognostic associations between GSRlncRNAs and BLCA. Based on these findings, we constructed a GSRlncRNA-based predictive signature, which demonstrated considerable potential for risk stratification and personalized treatment decisions. Functional experiments further validated the involvement of GSRlncRNAs in BLCA progression.